Granulocyte macrophage colony stimulating factor (GM-CSF; CSF2) is a well-studied protein which has long been appreciated for its hematopoietic properties (i.e. stimulation of proliferation and differentiation of progenitor cells and proliferation of mature cells of the myeloid lineage) (reviewed in Blood 77:1131, 1991; Rev Infect Dis 12: 41, 1990; Med. Oncol. 13:141, 1996). GM-CSF is constitutively produced by lung epithelial cells and the Paneth cells of the intestine (BBRC 312:897, 2003), but a wide variety of cells express GM-CSF upon activation with predominant expression from T cells, macrophages/monocytes, fibroblasts and endothelial cells (J Infect Dis 172:1573, 1995; J Infect Dis 185:1490, 2002; J Allergy Cin Immunol 112:653, 2003). The GM-CSF receptor (GM-CSFR; CSFR2) consists of a heterologous complex of two proteins; a high affinity alpha polypeptide which is specific for GM-CSF, and a low affinity common beta polypeptide which is shared by GM-CSF, IL-3 and IL-5 (reviewed in J Allergy Cin Immunol 112:653, 2003; Cytokine and Growth Factor Reviews 12:19, 2001). GM-CSFR is expressed on all cells of the myeloid lineage.
GM-CSF augments the activity of the innate immune system by mediating signals that cause or effect differentiation, survival, proliferation and activation of myeloid lineage cells including macrophages/monocytes, dendritic cells (DCs), neutrophils and eosinophils (reviewed in: J Immun 143:1198, 1989; Rev Infect Dis 12:41, 1990; Blood 77:1131, 1991; Trends in Immun. 23:403, 2002; Growth Factors 22:225, 2004). GM-CSF is an important factor for in vitro generation of monocyte-derived DCs and type 1 macrophages (PNAS 101:4560, 2004), and has been shown to induce differentiation and activation of DCs in vivo (Blood 95:2337, 2000). Human monocyte-derived macrophages generated in the presence of GM-CSF (Type 1 macrophages) produce high levels of proinflammatory cytokines such as IL-23, but not IL-12, whereas Type 2 macrophages generated in the presence of M-CSF (CSF1) produce anti-inflammatory cytokines such as IL-10, but not IL-23 (PNAS 101: 4560, 2004).
Human monocytes or macrophages stimulated with GM-CSF have increased function including cytotoxicity, production of other proinflammatory cytokines (IL-1β, TNFα and IL-6) and phagocytosis. Based on these effects, much effort has recently been applied to developing GM-CSF as a potent adjuvant for use in infectious disease or with administration of vaccines (reviewed in Eur J Clin Microbiol Infect Dis. 13::S47, 1994; Curr Opin Hematol. 7:168, 2000). Indeed, administration of rhGM-CSF in some clinical settings dramatically improves outcome and clearance of fungal infection (Eur J Clin Microbiol Infect Dis 13: S18, 1994; J Med Microbiol 47: 1998).
Microglia are the resident macrophages of the CNS and data from in vitro studies indicates that GM-CSF is a key cytokine which enhances survival, activation, proliferation and even differentiation of both fetal and adult microglial cells (Glia 12:309, 1994; J Immunol Methods 300:32, 2005). In addition, there are several reports from mouse MS model studies which provide evidence for a critical role of APCs (microglia or DCs) in the perivascular space of the CNS for disease initiation and persistence (Nat. Med. 11:146, 2005; Nat. Med. 11:328, 2005; Nat. Med. 11:335, 2005). GM-CSF stimulation of microglia upregulates MHCII and enhances antigen presentation
It is only recently that GM-CSF's role as a proinflammatory cytokine in disease, and dispensability as a hematopoietic growth factor, has been established (reviewed in Trends in Immun. 23: 403, 2002; Growth Factors 22: 225, 2004) and its role in causing or enhancing inflammatory/autoimmune disease.
Elevated levels of GM-CSF have been observed at local sites of inflammation in multiple sclerosis (MS), rheumatoid arthritis (RA), asthma, psoriasis, atopic dermatitis and sarcoidosis. Elevation of GM-CSF is not typically observed in the serum, thus determining disease association requires analysis of the target tissues. In MS, two clinical studies were performed in which levels of GM-CSF protein were measured by ELISA in cerebrospinal fluid (CSF) and serum from Relapsing-Remitting (RR) MS patients with active disease (new symptoms or worsening of existing symptoms within 2 weeks of tissue collection) and compared with either RRMS patients with stable disease (no episodes for prior 6 months) or other neurological disease (OND) controls (Eur Neurol 33:152, 1993; Immunopharmacol. Immunotoxicol. 20:373, 1998). Importantly, the OND controls did not include Alzheimer's Disease or vascular dementia patients, as highly increased levels of GM-CSF were reported in the CSF and sera of such patients (Acta Neurol Scand 103:166, 2001).
GM-CSF levels were in the low pg range, but were significantly higher in RRMS active disease CSF compared to stable disease CSF, and in MS active disease CSF compared to OND CSF. In addition, there were higher levels of TN F-alpha in CSF of active versus stable disease, and higher levels of both TGF-beta and IL-10 in CSF of stable versus active disease. The studies included very careful inclusion criteria with respect to ongoing treatment of patients and clinical definition of active versus stable disease, as well as synchronicity of sample collection. Interestingly, there were no significant differences in GM-CSF levels in serum between any of the groups. In addition, one study observed selective immunohistochemical detection of GM-CSF in astrocytes of MS lesions and not in control CNS white matter (n=3 MS donors, Glia 12:309, 1994). Finally, activated T cells and monocytes/macrophages are capable of producing large amounts of GM-CSF upon activation during an inflammatory response. There is ample evidence for the presence of both of these cell types in MS lesions (Ann Neurol. 47:707, 2000), and for T cells in CSF (reviewed in Curr. Neurol. Neurosci. Rep. 1: 257, 2001).
In addition to association of GM-CSF expression with MS, there is an abundance of disease association data for other inflammatory/autoimmune diseases and even some evidence for disease exacerbation with administration of exogenous GM-CSF. In RA, elevated levels of GM-CSF have been detected in synovial fluid (SF) of patients with RA or Psoriatic arthritis (PsA) compared to OA (bioassay, Clin. Exp. Immunol. 72:67, 1988) and compared to non-RA controls (bioassay, Rheumatol Int. 14:177, 1995). In addition, there is a strong correlation between the presence of CD68+ macrophages in joints with disease severity in RA patients (Ann Rheum Dis 64:834, 2005). Finally, it has been reported that GM-CSF treatment of RA patients with Felty's syndrome (neutropenia) can exacerbate disease (Blood 74:2769, 1989).
In asthma, GM-CSF has been found to be elevated in bronchial biopsies from asthmatic patients by immunohistochemistry and a correlation was observed between decrease in GM-CSF levels and increase in FEV1 following steroid treatment (Chest 105:687, 1994; Am Rev Respir Dis 147:1557, 1993). GM-CSF was also reported to be elevated in the sputum of intermittent, mild asthma patients (Ann Allergy Asthma Immunol 86:304, 2001). Data to support antagonism of GM-CSF includes a study in which the eosinophil promoting activity from BALF of symptomatic patients was attenuated by anti-GM-CSF mAb (in vitro, Eur. Respir. J. 12:872, 1998).
In psoriasis, GM-CSF expression was detected in psoriatic skin but not control skin samples (Arch Dermatol Res. 287:158, 1995; Clin Exp Dermatol. 19:383, 1994; Dermatologica. 181:16, 1990). It has also been reported that GM-CSF treatment of psoriasis can exacerbate disease (Br J Dermatol. 128:468, 1993).
In atopic dermatitis (AD), a significantly greater number of GM-CSF mRNA expressing cells were detected by in situ hybridization in biopsies of lesions of chronic AD than in acute AD or nonlesion skin (p<0.05; J Clin Invest. 95:211, 1995). In a second study, higher levels of GM-CSF were detected by immunohistochemistry of lesional AD skin (both epidermal and dermal compartments) and keratinocyte cultures established from uninvolved skin of AD patients exhibited increased spontaneous and PMA-stimulated production of GM-CSF compared with keratinocytes from nonatopic controls (J Clin Invest. 99:3009, 1997).
Mice deficient in GM-CSF (Science 264:713, 1994; PNAS 91:5592, 1994) and GM-CSFRc (Immunity 2:211, 1995; PNAS 92:9565, 1995) were generated by multiple groups. The mice had no overt differences in steady state levels of hematopoiesis, but did have histological evidence of alveolar proteinosis, were more susceptible to infections, and exhibited a modest delay in IgG production and diminished antigen-specific T cell responses after KLH immunization (PNAS 94:12557, 1997). GM-CSF−/− mice are resistant to MOG35-55-induced EAE (J Exp Med. 194:873, 2001), collagen-induced arthritis (CIA; JI 161:3639, 1998) and mBSA/IL-1-induced arthritis (Arthritis Rheum 44:111, 2001). In contrast, GM-CSF transgenic (Tg) mice have been generated in a number of labs and are associated with the development of inflammatory/autoimmune disease (Cell 51:675, 1987; JI 166:2090, 2001; J Clin Invest 97:1102, 1996; J Allergy Clin Immunol 111:1076, 2003; Lab Invest 77:615, 1997).
Thus there is a need in the art for GM-CSF inhibitors.